This proposal is for the development of algorithms to rapidly predict optimal beam modulation, number, and orientations for intensity modulated radiotherapy (IMRT). The innovation is the adaptation of tomographic techniques, suitably modified for radiotherapy dose and beams, to the inversion of a prescribed dose to an optimal beam configuration consistent with the available hardware. The proposed algorithms will assist the radiotherapy and radiosurgery planner in the efficient design of MeV photon therapy beams to conform dose to the tumor and avoid exposure of sensitive healthy tissue. The proposed system rapidly estimates a complex near-optimum beam configuration from among a large set of possible beam locations. This allows the clinician to obtain an initial estimate of an IMRT plan interactively. It is expected that the output beam modulation and configuration will be input to a full planning system for verification. The algorithms will be tested on various numerical and real-data tumor and organ-at-risk configurations. The objectives of the Phase I proposal are to develop IMRT software to optimize clinically achievable 1) beam number, 2) beam orientation, and 3) apply the algorithms to multi-isocenter radiosurgery of irregularly-shaped lesions. PROPOSED COMMERCIAL APPLICATION: The commercial application is a stand-alone commercial software module for IMRT inversion: including the prediction of optimum beam number and orientations. The applications include radiotherapy and multi-isometer radiosurgery planning.